Guidance for Industry Community-Acquired Bacterial Pneumonia: Developing Drugs for Treatment

Guidance for Industry Community-Acquired Bacterial Pneumonia: Developing Drugs for Treatment DRAFT GUIDANCE This guidance document is being distributed for comment purposes only. Comments and suggestions regarding this draft document should be submitted within 90 days of publication in the Federal Register of the notice announcing the availability of the draft guidance. Submit electronic comments to http://www.regulations.gov. Submit written comments to the Division of Dockets Management (HFA-305), Food and Drug Administration, 5630 Fishers Lane, rm. 1061, Rockville, MD 20852. All comments should be identified with the docket number listed in the notice of availability that publishes in the Federal Register. For questions regarding this draft document contact Sumathi Nambiar, MD, MPH or Joseph Toerner, MD, MPH at 301-796-1300. U.S. Department of Health and Human Services Food and Drug Administration Center for Drug Evaluation and Research (CDER) January 2014 Clinical/Antimicrobial Revision 2 10223dft.doc 01/08/14

Guidance for Industry Community-Acquired Bacterial Pneumonia: Developing Drugs for Treatment Additional copies are available from: Office of Communications, Division of Drug Information Center for Drug Evaluation and Research Food and Drug Administration 10903 New Hampshire Ave., Bldg. 51, rm. 2201 Silver Spring, MD 20993-0002 Tel: 301-796-3400; Fax: 301-847-8714; E-mail: druginfo@fda.hhs.gov http://www.fda.gov/drugs/guidancecomplianceregulatoryinformation/guidances/default.htm U.S. Department of Health and Human Services Food and Drug Administration Center for Drug Evaluation and Research (CDER) January 2014 Clinical/Antimicrobial Revision 2

TABLE OF CONTENTS I. INTRODUCTION... 1 II. III. BACKGROUND... 2 DEVELOPMENT PROGRAM... 3 A. General Considerations... 3 1. Nonclinical Development Considerations... 3 2. Drug Development Population... 3 3. Efficacy Considerations... 3 4. Safety Considerations... 3 B. Specific Efficacy Trial Considerations... 4 1. Trial Design... 4 2. Trial Population... 4 3. Entry Criteria... 4 a. Clinical, radiographic, and microbiologic entry criteria... 4 b. Exclusion criteria... 5 4. Randomization and Blinding... 5 5. Specific Populations... 6 6. Dose Selection... 6 7. Choice of Comparators, Prior Antibacterial Drug Use, and Concomitant Therapy... 6 8. Efficacy Endpoints... 8 a. Primary endpoint... 8 b. Secondary endpoints... 8 c. IV and oral formulations... 9 9. Trial Procedures and Timing of Assessments... 9 a. Entry visit... 9 b. On-therapy visits... 9 c. After therapy visit... 9 10. Statistical Considerations... 10 a. Analysis populations... 10 b. Noninferiority margins... 11 c. Sample size considerations... 11 11. Risk-Benefit Considerations... 12 C. Other Considerations... 12 1. Pharmacokinetic/Pharmacodynamic Evaluation... 12 2. Labeling Considerations... 13 REFERENCES... 14 APPENDIX: NONINFERIORITY MARGIN JUSTIFICATION FOR CABP... 16

1 Guidance for Industry 1 2 Community-Acquired Bacterial Pneumonia: 3 Developing Drugs for Treatment 4 5 6 7 8 This draft guidance, when finalized, will represent the Food and Drug Administration s (FDA s) current 9 thinking on this topic. It does not create or confer any rights for or on any person and does not operate to 10 bind FDA or the public. You can use an alternative approach if the approach satisfies the requirements of 11 the applicable statutes and regulations. If you want to discuss an alternative approach, contact the FDA 12 staff responsible for implementing this guidance. If you cannot identify the appropriate FDA staff, call 13 the appropriate number listed on the title page of this guidance. 14 15 16 17 18 I. INTRODUCTION 19 20 The purpose of this guidance is to assist sponsors in the clinical development of drugs for the 21 treatment of community-acquired bacterial pneumonia (CABP). Specifically, this guidance 22 addresses the Food and Drug Administration s (FDA s) current thinking regarding the overall 23 development program and clinical trial designs for drugs to support an indication for the 24 treatment of CABP. 2 This draft guidance is intended to serve as a focus for continued comments 25 and discussions among the Division of Anti-Infective Products, pharmaceutical sponsors, the 26 academic community, and the public. 3 27 28 This guidance does not contain discussion of the general issues of statistical analysis or clinical 29 trial design. Those topics are addressed in the ICH guidances for industry E9 Statistical 30 Principles for Clinical Trials and E10 Choice of Control Group and Related Issues in Clinical 31 Trials, respectively. 4 32 1 This guidance has been prepared by the Division of Anti-Infective Products in the Center for Drug Evaluation and Research (CDER) at the Food and Drug Administration. 2 For the purposes of this guidance, all references to drugs include both human drugs and therapeutic biological products regulated by CDER unless otherwise specified. 3 In addition to consulting guidances, sponsors are encouraged to contact the division to discuss specific issues that arise during drug development. 4 We update guidances periodically. To make sure you have the most recent version of a guidance, check the FDA Drugs guidance Web page at http://www.fda.gov/drugs/guidancecomplianceregulatoryinformation/guidances/default.htm. 1

33 This guidance revises the draft guidance for industry Community-Acquired Bacterial 34 Pneumonia: Developing Drugs for Treatment that issued in March 2009. When final, this 35 guidance will be considered the FDA s current thinking regarding the development of drugs for 36 the treatment of CABP. 37 38 FDA s guidance documents, including this guidance, do not establish legally enforceable 39 responsibilities. Instead, guidances describe the Agency s current thinking on a topic and should 40 be viewed only as recommendations, unless specific regulatory or statutory requirements are 41 cited. The use of the word should in Agency guidances means that something is suggested or 42 recommended, but not required. 43 44 45 II. BACKGROUND 46 47 This guidance provides information to assist sponsors developing drugs for the treatment of 48 CABP. CABP is defined as an acute bacterial infection of the pulmonary parenchyma associated 49 with chest pain, cough, sputum production, difficulty breathing, chills, rigors, fever, or 50 hypotension, and is accompanied by the presence of a new lobar or multilobar infiltrate on a 51 chest radiograph. Common typical bacterial pathogens that cause CABP include Streptococcus 52 pneumoniae, Haemophilus influenzae, Staphylococcus aureus, and Moraxella catarrhalis. 53 Atypical bacterial pathogens such as Chlamydophila pneumoniae, Mycoplasma pneumoniae, and 54 Legionella pneumophila also cause CABP. 55 56 Changes from the 2009 draft CABP guidance, based on public discussions and comments to the 57 docket, have been incorporated into the appropriate sections below. 5 These changes are intended 58 to attain a greater degree of balance between the practicability of conducting CABP clinical trials 59 and the trial procedures needed for a scientifically sound and interpretable trial. 60 61 5 There have been several public discussions with the FDA regarding CABP. For example: (1) a 2008 Clinical Infectious Diseases supplement that summarized a workshop co-sponsored by the FDA and professional societies, titled Workshop on Issues in the Design and Conduct of Clinical Trials of Antibacterial Drugs in the Treatment of Community-Acquired Pneumonia (Clinical Infectious Diseases, December 1, 2008; volume 47 (supplement number 3)); (2) a 2008 Anti-Infective Drugs Advisory Committee (AIDAC) meeting on endpoints and clinical trial design issues for CABP at http://www.fda.gov/ohrms/dockets/ac/cder08.html#antiinfective; (3) the December 9, 2009, AIDAC meeting on CABP issues at http://www.fda.gov/advisorycommittees/calendar/ucm187911.htm; and (4) the November 3, 2011, AIDAC meeting on CABP clinical trials at http://www.fda.gov/advisorycommittees/committeesmeetingmaterials/drugs/anti- InfectiveDrugsAdvisoryCommittee/ucm242307.htm (the November 3, 2011, AIDAC meeting information is at the bottom of the Web page). Notably, this revised guidance provides new efficacy endpoint recommendations (section III.B.8., Efficacy Endpoints), allows enrollment of up to 25 percent of the patient population who have received prior antibacterial drug therapy (section III.B.7., Choice of Comparators, Prior Antibacterial Drug Use, and Concomitant Therapy), and recommends the intent-to-treat population as the primary analysis population (section III.B.10., Statistical Considerations). 2

62 III. DEVELOPMENT PROGRAM 63 64 A. General Considerations 65 66 1. Nonclinical Development Considerations 67 68 New antibacterial drugs being studied for CABP should have nonclinical data documenting 69 activity against the commonly implicated pathogens for CABP. 70 71 2. Drug Development Population 72 73 The trial population should include individuals most likely to have CABP, as defined above, and 74 who can therefore benefit from antibacterial therapy. 75 76 3. Efficacy Considerations 77 78 Noninferiority trials are interpretable and acceptable to support approval of a drug for an 79 indication for the treatment of CABP. A showing of superiority to an effective control is also 80 readily interpretable and would be acceptable. 81 82 Historical data show that antibacterial drugs demonstrate a considerable treatment effect 83 compared to nonantibacterial therapies on clinical responses evaluated during the first 5 days of 84 therapy. 85 86 Although it remains important for a trial to demonstrate sustained clinical responses, currently 87 there is insufficient historical evidence to define the treatment effect on endpoints at or after 88 therapy completion. There is adequate information to define a reliable treatment effect on all 89 cause mortality. 90 91 A single adequate and well-controlled trial in CABP supported by evidence of antibacterial 92 activity accrued during a clinical development program (e.g., efficacy in another indication such 93 as acute bacterial skin and skin structure infection; data from a phase 2 clinical trial in CABP) 94 may provide evidence of effectiveness in CABP. Sponsors should discuss their proposed CABP 95 development program with the FDA as well as the other independent evidence that would be 96 used to support the findings from a single trial. 6 97 98 4. Safety Considerations 99 100 If the same or greater dose and duration of the drug is used in clinical development for other 101 infectious disease indications, safety data from the other infectious disease indications can be 102 used in an overall safety database to support an indication for CABP. In general, a minimum of 103 700 patients should be included in the safety database. For new drugs that have an important 104 clinical benefit over existing therapies, depending on the benefit demonstrated, a smaller 6 See the guidance for industry Providing Clinical Evidence of Effectiveness for Human Drug and Biological Products. 3

105 premarketing safety database may be appropriate. Sponsors should discuss the appropriate size 106 of the premarketing safety database with the FDA during clinical development. 107 108 B. Specific Efficacy Trial Considerations 109 110 1. Trial Design 111 112 CABP trials should be randomized, double-blind, and active-controlled using a noninferiority or 113 superiority design. Placebo-controlled trials are not appropriate for this indication except when 114 they are add-on superiority trials in which patients receive either placebo or investigational drug 115 added to standard-of-care antibacterial drug treatment. 116 117 2. Trial Population 118 119 The trial population for efficacy trials should include patients with CABP based on the entry 120 criteria described in section III.B.3., Entry Criteria. We recommend that at least 75 percent of 121 patients in trials have Pneumonia Patient Outcomes Research Team (PORT) scores of III or 122 higher (Fine, Auble, et al. 1997). For trials in which most patients would be treated as 123 outpatients, sponsors should discuss the trial population and its level of baseline severity with the 124 FDA in advance of a phase 3 trial (e.g., whether the trial may enroll patients with PORT scores 125 of II or higher). 126 127 3. Entry Criteria 128 129 a. Clinical, radiographic, and microbiologic entry criteria 130 131 Sponsors should use entry criteria that select patients who have evidence of a diagnosis of CABP 132 as outlined in Table 1. 133 134 Table 1. Summary of Entry Criteria for a CABP Trial At Least Two At Least Two At Least One Chest Microbiologic Symptoms Vital Sign Abnormalities Finding of Other Clinical Signs and Laboratory Abnormalities Radiograph Findings Criteria - Difficulty - Fever - Hypoxemia New Appropriate breathing - Hypotension - Clinical evidence of infiltrates in a sputum specimen: - Cough - Tachycardia pulmonary lobar or fewer than 10 - Production of - Tachypnea consolidation multilobar squamous purulent sputum - An elevated total distribution epithelial cells and - Chest pain white blood cell count or leukopenia more than 25 polymorphonuclear cells per low power field 135 136 An adequate specimen of respiratory secretions should be obtained in all patients and should be 137 processed by the laboratory according to recognized methods for Gram stain, culture, and in vitro 4

138 antibacterial susceptibility testing performed on appropriate organisms isolated from the 139 specimen. 7 140 141 Bacterial detection methods other than culture may be used to define the microbiological intent 142 to-treat (micro-itt) population (see section III.B.10.a., Analysis populations). Such methods 143 may include the following: (1) use of rapid diagnostic tests (e.g., urinary antigen test for S. 144 pneumoniae); and (2) nonculture methods of testing (e.g., serology, polymerase chain reaction). 145 Use of rapid diagnostic tests may help to select a patient population with an identified bacterial 146 etiology for CABP. 147 148 The clinical trial of an antibacterial drug also may provide an opportunity to contribute to the 149 development and evaluation of a new diagnostic test. Sponsors interested in also using a clinical 150 trial in patients with CABP as a means for the evaluation of a diagnostic test are encouraged to 151 discuss this with the FDA. 152 153 b. Exclusion criteria 154 155 Exclusion criteria should include the following: 156 157 Aspiration pneumonia 158 159 Hospital-acquired bacterial pneumonia or ventilator-associated bacterial pneumonia 160 161 Patients with known bronchial obstruction or a history of post-obstructive pneumonia 162 (this criterion does not exclude patients who have chronic obstructive pulmonary disease) 163 164 Patients with primary or metastatic lung cancer 165 166 Patients with cystic fibrosis, known or suspected Pneumocystis jiroveci pneumonia, or 167 known or suspected active tuberculosis 168 169 4. Randomization and Blinding 170 171 Patients should be randomized to treatment groups at enrollment. All trials should be double 172 blind unless there is a compelling reason for not blinding treatment allocation. If trials are 173 single-blind or open-label, sponsors should discuss potential biases with the FDA and how these 174 biases will be addressed. 175 7 Standard methods for in vitro susceptibility testing are developed by organizations such as the Clinical and Laboratory Standards Institute; see also the American Society for Microbiology, 2011, Manual of Clinical Microbiology, 10th edition. 5

176 5. Specific Populations 177 178 The trials should include patients of both sexes and all races, as well as geriatric patients. 8 179 Patients with renal or hepatic impairment may be enrolled, provided pharmacokinetics of the 180 drug have been evaluated in these patients and appropriate dosing regimens have been defined. 181 182 Sponsors should discuss drug development in the pediatric populations as early as is feasible. 183 The Pediatric Research Equity Act (PREA), as amended by the Food and Drug Administration 184 Safety and Innovation Act, states that initial plans for the conduct of pediatric studies (referred to 185 as an initial pediatric study plan) shall be submitted to the FDA before the date on which 186 required pediatric assessments are submitted under PREA and no later than (1) 60 days after the 187 end-of-phase 2 meeting or (2) such other time as may be agreed upon by the FDA and the 188 applicant. 9 189 190 6. Dose Selection 191 192 To choose the dose or doses to be evaluated in phase 3 clinical trials, sponsors should integrate 193 the findings from nonclinical toxicology studies, animal models of infection, pharmacokinetics, 194 safety and tolerability information from phase 1 clinical trials, and safety and efficacy 195 information from phase 2 dose-ranging clinical trials. Trials assessing drug penetration at the 196 site of action (e.g., epithelial lining fluid) may be helpful in defining doses that achieve 197 concentrations sufficient to exert an antibacterial effect. In addition, the pharmacokinetics of the 198 drug in specific populations (e.g., geriatric patients, patients with renal or hepatic impairment) 199 should be evaluated before initiation of phase 3 trials to determine whether dose adjustments are 200 necessary. This evaluation may prevent the exclusion of such patients from phase 3 clinical 201 trials. 202 203 7. Choice of Comparators, Prior Antibacterial Drug Use, and Concomitant Therapy 204 205 In general, the active comparator should be considered standard of care for this indication. 206 When evaluating the current standard of care, we consider recommendations by authoritative 207 scientific bodies (e.g., American Thoracic Society, Infectious Diseases Society of America) 208 based on clinical evidence and other reliable information that reflects current clinical practice. 209 210 Ideally, patients enrolled in a CABP clinical trial should not have received prior antibacterial 211 drug therapy because such therapy may have a number of potential consequences for a clinical 212 trial. Prior antibacterial drug therapy could: 213 8 See the ICH guidances for industry E7 Studies in Support of Special Populations: Geriatrics and E7 Studies in Support of Special Populations: Geriatrics; Questions and Answers. 9 See PREA (Public Law 108-155; section 505B of the Federal Food, Drug, and Cosmetic Act; 21 U.S.C. 355c) as amended by the Food and Drug Administration Safety and Innovation Act of 2012 (Public Law 112-144) and the draft guidance for industry Pediatric Study Plans: Content of and Process for Submitting Initial Pediatric Study Plans and Amended Pediatric Study Plans. When final, this guidance will represent the FDA s current thinking on this topic. 6

214 Obscure true treatment differences between an investigational drug and the control drug 215 introducing bias toward a finding of no difference between treatment groups (i.e., a bias 216 toward noninferiority) 10 217 218 Particularly influence the efficacy findings based on an endpoint early in therapy (day 3 219 to day 5) 220 221 However, exclusion of all patients who have received prior antibacterial therapy also may pose 222 problems, including: 223 224 Excluding patients with greater disease severity (i.e., patients who received prompt 225 administration of antibacterial drug therapy), which may result in a patient population 226 with lesser severity of illness and greater potential for spontaneous recovery; this could 227 bias trial results toward a finding of no difference between treatment groups (i.e., a bias 228 toward noninferiority) 229 230 Certain trial sites may not participate in the clinical trial because of concerns that trial 231 treatment would not represent standard of care. 232 233 A pragmatic approach to these concerns is to: (1) encourage prompt enrollment procedures so 234 that patients can receive the clinical trial treatment initially, with no need for other antibacterial 235 drug therapy; and (2) allow enrollment of some patients who have received a single dose of a 236 short-acting antibacterial drug within 24 hours of enrollment (ideally there would be few such 237 patients but up to 25 percent of the patient population could be allowed). This would permit 238 patients in the trial to receive prompt antibacterial drug therapy as clinically necessary, consistent 239 with the standard of care. The results in the subgroup of patients (i.e., the majority of patients) 240 who did not receive prior effective antibacterial drug therapy would be important to evaluate. 241 The primary analysis should be stratified by prior therapy to assess the consistency of the results 242 across the two subgroups (i.e., patients who received prior therapy and those who did not receive 243 prior therapy). 244 245 In general, concomitant antibacterial therapy with an antimicrobial spectrum that overlaps with 246 the spectrum of the investigational drug should not be administered during the trial. We 247 recognize the need in certain circumstances for the empirical coverage against atypical pathogens 248 (e.g., Legionella species). The additional antibacterial coverage for atypical pathogens should be 249 discussed with the FDA before trial initiation. The additional antibacterial coverage for atypical 250 pathogens should be promptly discontinued after a determination has been made that CABP is 251 not caused by an atypical pathogen of concern (e.g., a negative test result on a Legionella antigen 252 assay). 253 10 For example, see Pertel, Bernardo, et al. 2008. 7

254 8. Efficacy Endpoints 255 256 a. Primary endpoint 257 258 The primary efficacy endpoint of clinical success is defined as improvement at day 3 to day 5 in 259 at least two of the following symptoms: chest pain, frequency or severity of cough, amount of 260 productive sputum, and difficulty breathing. 11 Symptoms should be evaluated on a four-point 261 scale (absent, mild, moderate, severe) with improvement defined as at least a one-point 262 improvement from baseline to the assessment at day 3 to day 5 (e.g., from severe to moderate, 263 from moderate to absent, or from mild to absent). 12 264 265 An endpoint of all-cause mortality at 28 days after enrollment may be used as a primary efficacy 266 endpoint in CABP clinical trials in certain patient populations. However, sponsors considering 267 the use of all-cause mortality as the primary efficacy endpoint should discuss the trial design 268 with the FDA. 269 270 b. Secondary endpoints 271 272 Sponsors should evaluate the following as secondary endpoints: 273 274 Improvement at day 3 to day 5 in at least two of the following symptoms with no 275 worsening in any of these symptoms of CABP compared to baseline: chest pain, 276 frequency or severity of cough, amount of productive sputum, and difficulty breathing; 277 and improvement in vital signs (i.e., body temperature, blood pressure, heart rate, 278 respiratory rate). 13 279 280 Clinical outcome at the end of therapy. 281 282 Clinical outcome at a fixed time point after therapy completion. Patients with resolution 283 of symptoms and signs attributable to CABP at 5 to 10 days following completion of 284 treatment and who did not receive nontrial antibacterial drugs for treatment of CABP 285 should be considered successes on this secondary endpoint. 286 287 Other examples of secondary endpoints for consideration are as follows: 288 289 Changes in white blood cell counts from baseline to day 3 to day 5 290 Changes in oxygenation from baseline to day 3 to day 5 291 11 See Talbot, Powers, et al. 2012. 12 See Toerner, Burke, et al. 2012. For information regarding the development of patient-reported outcome measures, see the guidance for industry Patient-Reported Outcome Measures: Use in Medical Product Development to Support Labeling Claims. 13 Improvement or stabilization of vital signs and other signs attributable to CABP should be defined in the protocol. For example, see table 10 in Mandel, Wunderink, et al. 2007. 8

292 c. IV and oral formulations 293 294 For drugs that have only an intravenous (IV) formulation available, sponsors should conduct 295 trials with the IV formulation alone until the day 3 to day 5 efficacy endpoint assessment is 296 complete, if feasible, to allow for assessment of both the efficacy and safety of the 297 investigational drug. Assessment of the primary endpoint at day 3 to day 5 before switching to 298 an oral antibacterial drug should ensure that the evaluation of efficacy reflects the effects of the 299 investigational IV drug. The overall duration of antibacterial drug therapy (i.e., days of IV 300 therapy plus days of oral drug therapy) should not involve an unnecessarily long course of oral 301 switch therapy, so that the contribution of the IV investigational drug to overall efficacy on 302 secondary endpoints at 5 to 10 days after completion of treatment can be assessed. 303 304 For drugs that have both an IV and oral formulation, the protocol should specify the criteria that 305 allow for IV-to-oral switch. The sponsor should collect pharmacokinetic (PK) data for IV and 306 oral formulations in earlier phase studies to select the appropriate oral dose for the IV-to-oral 307 switch. 308 309 9. Trial Procedures and Timing of Assessments 310 311 a. Entry visit 312 313 The following information should be captured at the entry visit (see section III.B.3., Entry 314 Criteria, and section III.B.8., Efficacy Endpoints): 315 316 Appropriate demographic information 317 History and physical examination findings 318 Prior medication use 319 Baseline assessments of symptoms 320 Baseline assessments of clinical signs of CABP 321 Baseline appropriate laboratory tests 322 Chest radiographic findings 323 Microbiological specimens 324 Severity scores 325 326 b. On-therapy visits 327 328 Investigators should document findings from on-therapy clinical trial visits (e.g., history, 329 physical examination, adverse effects, laboratory test results). Patients should be evaluated for 330 the symptoms of chest pain, frequency or severity of cough, amount of productive sputum, and 331 difficulty breathing at day 3 to day 5. Patients also should be evaluated at the end of therapy. 332 333 c. After therapy visit 334 335 At this visit at 5 to 10 days after completion of treatment, sponsors should capture physical 336 examination findings, assessments of symptoms, assessments of signs, assessments and 9

337 resolution of adverse effects, if any, and appropriate laboratory tests. Patients should be 338 evaluated at day 28 for assessment of all-cause mortality. 339 340 10. Statistical Considerations 341 342 The trial hypotheses and the analysis methods should be prespecified in the protocol and in the 343 statistical analysis plan, and should be finalized before trial initiation. 14 344 345 a. Analysis populations 346 347 The following definitions apply to various analysis populations in CABP clinical trials: 348 349 Safety population All patients who received at least one dose of drug during the trial. 350 351 Intent-to-treat (ITT) population All patients who were randomized. 352 353 Micro-ITT population All randomized patients who have a baseline bacterial pathogen 354 known to cause CABP against which the investigational drug has antibacterial activity. 355 This includes bacterial pathogens identified by standard culture methods of an 356 appropriate sputum specimen or blood. Recently conducted trials suggest that 357 approximately 25 percent of the ITT population will have bacterial pathogens identified 358 by standard culture methods. In addition, nonculture methods of detection of bacterial 359 pathogens (e.g., urinary antigen test) may be used to identify patients for inclusion in a 360 micro-itt analysis population. 361 362 Clinically evaluable or per-protocol populations Patients who meet the definition for 363 the ITT population and who follow important components of the trial as specified in the 364 protocol. 365 366 Microbiologically evaluable populations Patients who meet the definition for the 367 micro-itt population and who follow important components of the trial as specified in 368 the protocol. 369 370 Sponsors should discuss with the FDA the prespecified primary analysis population in advance 371 of trial initiation. The ITT population may be considered as the primary analysis population 372 when (1) the trial enrolls patients who are most likely to have a bacterial etiology for pneumonia 373 and (2) the investigational antibacterial drug can be administered as monotherapy that has 374 antibacterial activity against the typical bacterial pathogens that cause CABP. 15 375 376 The ITT population is likely to have a substantial fraction of patients who do not have a bacterial 377 pathogen identified on sputum culture. Nonetheless, the ITT population (i.e., patients who meet 14 See ICH E9 and ICH E10, and the draft guidance for industry Non-Inferiority Clinical Trials (when final, this guidance will represent the FDA s current thinking on this topic). 15 The micro-itt population is an important analysis population, in particular if the investigational antibacterial drug has a narrow spectrum of activity (e.g., a drug active against a single genus and species of bacteria). 10

378 the inclusion criteria described in section III.B.3, Entry Criteria) may be informative based on 379 observations from previously conducted trials and evaluations. For instance, among patients 380 who did not receive prior therapy in a trial in which there was an observed treatment difference 381 between two antibacterial drugs (Pertel, Bernardo, et al. 2008), the subgroup of patients who did 382 not have a positive sputum culture for a bacterial pathogen showed a treatment difference similar 383 to the treatment difference among the subgroup of patients with a positive culture. This indicates 384 a strong likelihood that the patients enrolled in this trial without a positive sputum culture 385 actually had bacterial disease (Rubin, Toerner, et al. 2012). In addition, extensive nonculture 386 methods performed in a research setting from sputum specimens identified a possible bacterial 387 etiology for pneumonia in some patients who did not have a bacterial pathogen identified on a 388 sputum or blood culture (Johansson, Kalin, et al. 2010). Another evaluation of patients with 389 pneumonia who did not have a bacterial pathogen identified on a sputum or blood culture found 390 that a more invasive search can identify a bacterial etiology in a large proportion of patients 391 (Ruiz-González, Falguera, et al. 1999). 392 393 However, sponsors planning to develop a drug for the sole indication of the treatment of CABP 394 should consider conducting two adequate and well-controlled trials of identical design. Each of 395 these trials could potentially be powered based on the ITT population of that trial. Further, a 396 noninferiority efficacy analysis in a micro-itt population could potentially use data pooled from 397 both trials. Sponsors planning to conduct a single CABP trial, with other supportive data, to 398 support approval for CABP should discuss this plan with the FDA in advance and are 399 encouraged to submit a special protocol assessment. 16 400 401 The micro-itt population should allow a sufficient description of baseline microbiological 402 findings for adequate labeling information. 403 404 b. Noninferiority margins 405 406 Historical experience indicates that there is a relatively large treatment effect of antibacterial 407 therapy on clinical recovery at day 3 to day 5 (see the Appendix). In general, the selection of a 408 noninferiority margin (M 2 ) of 12.5 percent is reasonable for CABP clinical trials using a clinical 409 recovery endpoint at day 3 to day 5. In certain circumstances (e.g., a narrow spectrum drug for a 410 limited population with unmet medical need), it may be reasonable to consider a noninferiority 411 margin greater than 12.5 percent. Sponsors should discuss with the FDA a clinically appropriate 412 noninferiority margin in advance of trial initiation. 413 414 c. Sample size considerations 415 416 A general framework is provided for sponsors to begin to discuss sample size considerations 417 with the FDA during protocol development. In this illustrative sample size calculation, 418 approximately 225 patients per group is estimated based on the following assumptions: (1) a rate 419 of clinical success for the active-controlled therapy of 80 percent; (2) two-sided type I error (α) 420 of 0.05; (3) type II error (β) of 0.10 (power 0.90); (4) a noninferiority margin of 12.5 percent (see 421 the Appendix); and (5) an ITT analysis population. 422 16 See the guidance for industry Special Protocol Assessment. 11

423 11. Risk-Benefit Considerations 424 425 Risk-benefit considerations may depend on the population being studied. For example, for an 426 IV-administered antibacterial drug targeted for treatment of hospitalized patients seriously ill 427 with CABP, certain types of adverse effects that can be monitored in a hospital setting might 428 result in a risk-benefit consideration that is appropriate, while such adverse effects might result 429 in a risk-benefit consideration that is not appropriate for an orally administered antibacterial drug 430 targeted for treatment of mildly ill outpatients. 431 432 C. Other Considerations 433 434 1. Pharmacokinetic/Pharmacodynamic Evaluation 435 436 The PK/pharmacodynamic (PD) characteristics of the drug should be evaluated using in vitro 437 methods and animal models of infection. 438 439 The limitations of S. pneumoniae pneumonia and H. influenzae pneumonia animal models, when 440 considering their implications for humans, include the differences among the animal models in 441 the mode of infection and in the reproducibility of infection (Tessier, Kim, et al. 2002; Gavalda, 442 Capdevila, et al. 1997; Legget 1999; Miyazaki, Nunoya, et al. 1997), and differences in the effect 443 of animal lung secretions versus human lung secretions on the activity of the antibacterial drug 444 (Silverman, Mortin, et al. 2005). Animal studies are not a substitute for clinical trials in patients 445 with CABP. 17 446 447 The PK/PD characteristics of the drug (including the relationships to the minimum inhibitory 448 concentrations) should be integrated with the findings from phase 1 PK clinical trials to help 449 identify appropriate dosing regimens for evaluation in phase 2 and phase 3 clinical trials. A 450 dose-response trial may be considered as an option for clinical trials early in development to 451 weigh risks and benefits when selecting doses and to ensure that suboptimal doses or excessive 452 doses (beyond those that add to efficacy) are not used in the phase 3 trial, offering some 453 protection against unexpected and unrecognized dose-related toxicity. 18 454 455 Sponsors should consider obtaining blood samples from all patients in phase 2 and phase 3 456 clinical trials (sparse sampling) to allow for the estimation of drug exposure in each patient. A 457 retrospective exposure-response analysis based on the population PK model should be performed 458 to assess the relationship between exposure and observed clinical and microbiologic outcomes. 459 The relationship between drug exposure and clinically relevant adverse events also should be 460 explored to identify potential risks with different dosing regimens (if applicable) and specific 461 patient populations. 462 17 See 21 CFR 314.600 (http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/cfrsearch.cfm?fr=314.600) 18 See the guidance for industry Exposure-Response Relationships Study Design, Data Analysis, and Regulatory Applications and the ICH guidance for industry E4 Dose-Response Information to Support Drug Registration. 12

463 2. Labeling Considerations 464 465 Generally, the labeled indication should be the treatment of CABP caused by the specific 466 bacteria identified in a sufficient number of patients in the clinical trials and should reflect the 467 patient population enrolled in the clinical trials. 468 13

469 REFERENCES 470 471 Austrian, R and J Gold, 1964, Pneumococcal Bacteremia With Especial Reference to Bacteremic 472 Pneumococcal Pneumonia, Ann Intern Med, 60:759-776. 473 474 Bullowa, JGW, 1937, The Course, Symptoms and Physical Findings, In: Bullowa JGW, editor, 475 The Management of Pneumonias, Oxford University Press; New York. 476 477 Dowling, HG and MH Lepper, 1951, The Effect of Antibiotics (Penicillin, Aureomycin and 478 Terramycin) on the Fatality Rate and Incidence of Complications in Pneumococcic Pneumonia: 479 A Comparison With Other Methods of Therapy, AM J Med Sci, 222:396-402. 480 481 Fine, MJ, TE Auble, DM Yealy, BH Hanusa, LA Weissfeld, DE Singer, CM Coley, TJ Marrie, 482 and WN Kapoor, 1997, A Prediction Rule to Identify Low-Risk Patients With Community 483 Acquired Pneumonia, N Engl J Med, 336:243-50. 484 485 Finland, M, 1943, Chemotherapy in the Bacteremia, Conn State Med J, 7:92-100. 486 487 Finland, M, WC Spring, and FC Lowell, 1940, Specific Treatment of the Pneumococcic 488 Pneumonias; An Analysis of the Results of Serum Therapy and Chemotherapy at the Boston City 489 Hospital From July 1938 Through June 1939, Annals of Internal Medicine, 13:1567-1593. 490 491 Flippin, HF, JS Lockwood, DS Pepper, and L Schwartz, 1939, The Treatment of Pneumococcic 492 Pneumonia With Sulfapyridine, JAMA, 112:529-534. 493 494 Gavaldà, J, JA Capdevila, B Almirante et al., 1997, Treatment of Experimental Pneumonia due 495 to Penicillin-Resistant Streptococcus Pneumoniae in Immunocompetent Rats, Antimicrob Agents 496 Chemother, 41:795-801. 497 498 Higgins, K, M Singer, T Valappil, S Nambiar, D Lin, and E Cox, 2008, Overview of Recent 499 Studies of Community-Acquired Pneumonia, Clin Infect Dis, 47 (Suppl 3) S150-S156. 500 501 Johansson, N, M Kalin, A Tivelijung-Lindell, CG Giske, and J Hedlund, 2010, Etiology of 502 Community-Acquired Pneumonia: Increased Microbial Yield With New Diagnostic Methods, 503 Clin Infect Dis, 50:202-209. 504 505 Kingston, JR, RM Chanock, MA Mufson et al., 1961, Eaton Agent Pneumonia, JAMA, 176:118 506 123. 507 508 Legget, J, 1999, Murine Models of Pneumonia Using Aerosol Infection, In: Zak O, Sande MA, 509 eds., Handbook of Animal Infections: San Diego, Academic Press, 533-538. 510 511 Mandell, LA, RG Wunderink, A Anzueto et al., 2007, Infectious Diseases Society of 512 America/American Thoracic Society Consensus Guidelines on the Management of Community 513 Acquired Pneumonia in Adults, Clin Infect Dis, 44:S27-72. 514 14

515 Meakins, JC and FR Hanson, 1939, The Treatment of Pneumoccic Pneumonia With 516 Sulfapyridine, The Canadian Medical Association Journal, April, 333-336. 517 518 Miyazaki, S, T Nunoya, T Matsumoto, K Tateda, and K Yamaguchi, 1997, New Murine Model 519 of Bronchopneumonia due to Cell-Bound Haemophilus Influenzae, J Infect Dis, 175:205-209. 520 521 Pertel, PE, P Bernardo, C Fogarty et al., 2008, Effects of Prior Effective Therapy on the Efficacy 522 of Daptomycin and Ceftriaxone for the Treatment of Community-Acquired Pneumonia, Clin 523 Infect Dis, 46:1142-1151. 524 525 Rubin, D, J Toerner, T Valappil et al., 2012, Impact of Prior Antibacterial Therapy in 526 Community-Acquired Bacterial Pneumonia (CABP) Trials, Infectious Diseases Society of 527 America, October 17-21, Abstract number 36677. 528 529 Ruiz-González, A, M Falguera, A Nogués, and M Rubio-Caballero, 1999, Is Streptococcus 530 Pneumoniae the Leading Cause of Pneumonia of Unknown Etiology? A Microbiologic Study of 531 Lung Aspirates in Consecutive Patients With Community-Acquired Pneumonia, Am J Med., 532 Apr, 106(4):385-90. 533 534 Silverman, JA, LI Mortin, AD Vanpraagh, T Li, and J Alder, 2005, Inhibition of Daptomycin by 535 Pulmonary Surfactant: In Vitro Modeling and Clinical Impact, J Infect Dis, 191(12):2149-2152. 536 537 Singer, M, S Nambiar, T Valappil, K Higgins, and S Gitterman, 2008, Historical and Regulatory 538 Perspectives on the Treatment Effect of Antibacterial Drugs for Community-Acquired 539 Pneumonia, Clin Infect Dis, 47 (Suppl 3): S216-S224. 540 541 Talbot, GH, JH Powers, TR Fleming et al., 2012, Progress on Developing Endpoints for 542 Registrational Clinical Trials of Community-Acquired Pneumonia and Acute Bacterial Skin and 543 Skin Structure Infections: Update From the Biomarkers Consortium of the Foundation for the 544 National Institutes of Health, Clin Infect Dis, 55:1114-1121. 545 546 Tessier, PR, MK Kim, W Zhou et al., 2002, Pharmacodynamic Assessment of Clarithromycin in 547 a Murine Model of Pneumococcal Pneumonia, Antimicrob Agents Chemother, 46:1425-1434. 548 549 Toerner, JG, L Burke, S Komo, E Papadopoulos, 2012, A Collaborative Model for Endpoint 550 Development for Acute Bacterial Skin and Skin Structure Infections and Community-Acquired 551 Bacterial Pneumonia, Clin Infect Dis, 55:1122-1123. 552 553 Wilson, AT, AH Spreen, ML Cooper et al., 1939, Sulfapyridine in the Treatment of Pneumonia 554 in Infancy and Childhood, JAMA, 112:1435-1439. 15

555 APPENDIX: 556 NONINFERIORITY MARGIN JUSTIFICATION FOR CABP 557 558 Background 559 560 The selection of a noninferiority margin depends on a reliable estimate of the treatment effect of 561 the active comparator (i.e., effect of the active comparator over placebo, referred to as M 1 ), 562 usually based upon placebo-controlled trials, that can be assumed to hold for the noninferiority 563 trial. After M 1 is established, clinical judgment determines how much of the estimated treatment 564 effect (M 1 ) should be preserved in determining a clinically acceptable noninferiority margin, 565 referred to as M 2. 566 567 Historical studies and clinical trials of antibacterial treatment of bacterial pneumonia provide 568 evidence that antibacterial drugs have the following effects: 569 570 Achievement of a greater proportion of patients with favorable clinical responses at time 571 points earlier in the course of antibacterial drug therapy (i.e., at day 3 to day 5) 572 573 Reduction of mortality in patients with pneumococcal or lobar pneumonia 574 575 An area of uncertainty in evaluating historical data is the spectrum of bacterial pathogens that 576 cause CABP today. In most of the historical studies and historical-controlled clinical trials, 577 CABP was considered synonymous with pneumococcal pneumonia because S. pneumoniae was 578 regularly identified. A review of recently conducted trials showed that less than 20 percent of 579 the total patient populations had documented S. pneumoniae (Higgins, Singer, et al. 2008). 580 CABP is also caused by other pathogens such as H. influenzae, H. parainfluenzae, S. aureus, and 581 M. catarrhalis, as well as atypical bacteria such as M. pneumoniae, C. pneumoniae, and 582 Legionella species. Limited information is available on antibacterial treatment effect in CABP 583 caused by M. pneumoniae (Kingston, Chanock, et al. 1961). A fundamental assumption is that 584 historical response rates in infections such as S. pneumoniae CABP are relevant to response rates 585 in modern infections with sensitive organisms. 586 587 We describe the steps taken to determine a noninferiority margin for two primary outcome 588 measures: (1) an endpoint based on the outcome assessments of chest pain, frequency or severity 589 of cough, amount of productive sputum, and difficulty breathing; and (2) all-cause mortality 590 endpoint. 591 592 1. Endpoint Based on Clinical Outcome Assessments at Day 3 to Day 5 After Enrollment 593 594 Studies conducted around the time of the introduction of antibacterial drug therapy described 595 clinical responses among untreated patients and patients treated with antibacterial drugs. These 596 observational studies provide an estimate of the effect of antibacterial drugs on clinical response 597 endpoints other than mortality. 598 599 Several papers described the clinical course of patients with pneumococcal pneumonia in a 600 similar way; patients were recorded as having a successful clinical result by the demonstration of 16

601 fever resolution and accompanying improvement and resolution of other signs and symptoms of 602 pneumonia. For example, a description in one of the papers stated, This fall in temperature was 603 in all cases accompanied by a conspicuous reduction in the pulse and respiratory rates, and the 604 patients were improved subjectively (Meakins and Hanson 1939). One study described the 605 clinical course of 663 patients who did not receive antibacterial drug therapy (Bullowa 1937), 606 while two other studies included patients who received antibacterial drug therapy. One study 607 described the clinical course in 100 patients with pneumococcal pneumonia (Flippin, Lockwood, 608 et al. 1939) and another study described the clinical course in 30 patients with pneumococcal 609 pneumonia (Meakins and Hanson 1939). Figure A compares the three studies in the rates of 610 clinical recovery, defined generally as the improvement in both clinical signs and symptoms. 611 612 Figure A. Rates of Clinical Recovery Recorded at Each Day 100% Sample sizes: Bullowa: N = 663; Flippin: N = 100; Meakins: N = 30 90% Untreated (Bullowa) Sulfapyridine (Flippin) Sulfapyridine (Meakins) 80% 70% Recovery (%) 60% 50% 40% 30% 20% 10% 0% 1 2 3 4 5 6 7 613 614 615 The difference in clinical recovery rates between patients in the two treatment studies and 616 patients in the study without treatment were 72 percent and 77 percent. 617 618 Figure B shows the rates of clinical recovery in an observational study of patients with 619 pneumococcal pneumonia who received antibacterial drug therapy (sulfapyridine) and a group of 620 patients who received no specific therapy. Clinical recovery was defined as permanent drop in Days 17

621 oral temperature below 100 F, with subsidence of other symptoms of acute infection (Finland, 622 Spring, et al. 1940). Time points at 36 to 48 hours and 48 to 72 hours after therapy initiation 623 demonstrate the greatest treatment effect of clinical recovery. The treatment difference is 624 approximately 30 percent (95 percent confidence interval: 22 percent, 37 percent) at the 48- to 625 72-hour time point. Clinical observations that were reported at any time after the 48- to 72-hour 626 assessment are displayed as 72+ in Figure B. The time points after 72 hours (i.e., 72+) included 627 recovery time points out to several weeks following therapy completion. 628 629 Figure B. Rates of Clinical Recovery of Acute Bacterial Pneumonia (Finland, Spring, et al. 630 1940) 90% 80% Sulfapyridine (N=225) No Therapy (N=472) 71% 79% 70% 63% 69% 60% 55% Recovery (%) 50% 40% 33% 41% 30% 28% 31% 20% 16% 21% 10% 12% 0% 12 12--24 24--36 36--48 48--72 72+ Hours 631 632 Another paper described the outcomes among pediatric patients with pneumococcal pneumonia 633 and provides additional support for a treatment effect of antibacterial drugs relatively early in 634 therapy. The mean time to clinical recovery was 4.7 days among patients who received 635 antibacterial drug therapy while patients who did not receive antibacterial drug therapy had a 636 mean time to clinical recovery of 8.9 days (Wilson, Spreen, et al. 1939). 637 638 The clinical response endpoints that were evaluated in each of these studies were not well 639 defined. The studies evaluated both signs and symptoms together. A large treatment effect was 640 observed at the early time point in the course of therapy (i.e., day 3 to day 5 after therapy 18

641 initiation) for an endpoint that included improvement in both signs and symptoms. The studies 642 show that the treatment differences become smaller at times beyond day 3 to day 5 of therapy. 643 Aspects that support the use of these studies as an estimate of M 1 include the following: 644 645 The studies documented bacterial pneumonia, all as S. pneumoniae. 646 647 The estimate of the treatment difference appears to be large and is consistent across 648 studies. 649 650 Some patients included in the no therapy group in Figure B were patients who had signs 651 and symptoms of milder pneumonia. Even after the availability of antibacterial drugs, 652 the clinician chose not to treat such patients with antibacterial drug therapy because of the 653 likelihood of spontaneous recovery. The inclusion of patients more likely to experience 654 spontaneous recovery of pneumonia in the no therapy group leads to an underestimate of 655 the true treatment difference among patients with more serious disease. 656 657 The clinical response measurements are plausible consequences of treating an infection. 658 659 The limitations of these studies include the following: 660 661 The studies were not randomized 662 663 Historically controlled studies create a greater level of uncertainty in the estimate of 664 treatment differences 665 666 The clinical response evaluations were not defined 667 668 The clinical response evaluations included improvement in both signs and symptoms 669 together and did not separately evaluate improvement in chest pain, frequency or severity 670 of cough, amount of productive sputum, and difficulty breathing 671 672 The treatment difference appears to be large for an endpoint based on clinical outcome 673 assessments earlier in the course of therapy for CABP. However, the results are variable, 674 ranging from the point estimate of 30 percent treatment difference at a 48- to 72-hour time point 675 noted in Figure B to a point estimate of 77 percent treatment difference at day 3 noted in Figure 676 A. 677 678 It is difficult to provide a precise numerical value for the treatment effect of a proposed primary 679 endpoint of symptom improvement at day 3 to day 5. However, an M 1 of at least 20 percent 680 appears to be a reasonably appropriate and conservative estimate, accounting for the 681 uncertainties with clinical recovery in the historical literature. A conservative estimate of M 1 at 682 20 percent is still large enough to support the selection of a noninferiority margin (M 2 ) of 12.5 683 percent for the endpoint of symptom improvement at day 3 to day 5. The selection of the 684 noninferiority margin (M 2 ) is a matter of clinical judgment and should be justified by the 685 sponsor. 686 19